JP2885544B2 - Dead time compensation controller - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an improvement of a control device using a dead time Smith compensation method.

[0002]

2. Description of the Related Art Controllers having PI or PID control functions have been widely used in all industrial fields since the history of process control, and are no longer indispensable for plant operation.

By the way, if the controlled object is approximated by the dead time L and the first-order delay T (time constant), when the controlled object is only the first-order delay, it can be easily controlled by PID control.
When the dead time L is included, as the dead time L increases, that is, as L / T increases, the control becomes more difficult only with the PID control.

Therefore, as a method of improving the controllability of a control system including a dead time, O.S. J. MSmit (person name)
Is a so-called Smith in which a dead time compensator using a controlled object model is added to the PID control, and the dead time is taken out of the control loop to control only the apparently first order controlled object. Or dead time Smith compensation methods have been proposed and are now widely used.

FIG. 2A is a diagram showing functional blocks of a control device using the dead time Smith compensation method. The apparatus includes a deviation calculating means 1 for determining the error En between the target value SV _n and the controlled variable PV _n, on the basis of the deviation En PI
Alternatively, a PID adjustment operation is performed, and the obtained operation signal MV is obtained.
_This is a control system including a PID adjusting unit 3 for applying _n to the control target 2 and a dead time compensating unit 5 newly added.

[0006] The dead time compensator 5, the model unit 6 lag primary outputs an operation signal MV _n of PID adjustment device 3 through a primary delay transfer function, likewise the operation signal MV _n the first-order lag of the PID regulating means 3 And a controlled object model means (also referred to as a controlled object model) 7 for outputting through a transfer function having a dead time, a subtraction means 8 for subtracting the output of the controlled object model means 7 from the output of the primary delay model means 6, and the like. The output of the subtraction means 8 is introduced into the subtraction means 4 provided on the output side of the deviation calculation means 1, and the output of the dead time compensator 5 is subtracted from the deviation En. However, in FIG. 2A, _GP e ^{-Lp s} : The control target transfer function _{G P = K p / (1} + T P s): the control target of the transfer function without the dead time from the characteristic L _p: dead time K _p of the controlled object: a control target of the gain T _p: control object S: Laplace operator G _M · e ^-LMs : Transfer function of the controlled object model G _M = K _M / (1 + T _M s): Transfer function excluding the dead time from the controlled object model L _M : Control dead time of the object model K _M: the controlled object model gain T _M: is represented by the time constant of the controlled object model and rearranging the FIGS. 2 (a) equivalent transformation to become as shown in FIG. 2 (b) . Here, the condition that the disturbance D is small and can be ignored, and the characteristics of the controlled object and the characteristics of the controlled object model 7 match, that is,

[0007] The disturbance = small, and T _p = T _M, is assumed to be in the _{L p = L M ...... (1} ) the relationship, G _P = G _M, and the transfer function between the SV _n → PV _n in this case When _{seeking, PV n / SV n = {} (G c · G M) / (1 + G c · G M)} e -LP ^s (2) and can be converted into a configuration as shown in FIG. Therefore, this control device means that the first-order lag model means 6 excluding the dead time should be feedback-controlled by the PID adjusting means 3. In other words, this device can eliminate dead time from the control loop,
It can be easily controlled by the D adjusting means 3, and good controllability can be expected. Note that the dead time element 9 is output outside the control loop.

[0008]

However, as apparent from the above description, the control device using the dead time Smith compensation method as described above requires the condition shown in FIG. It cannot be configured as shown in FIG.

However, in actual plant control,
It is difficult to always satisfy the condition of the expression (1). For example, the condition of the expression (1) shifts with time due to a change in the characteristics of the controlled object 2 or an environmental change such as an ambient temperature, a catalyst concentration, a raw material condition, and a load. Things. As a result, the more the condition of equation (1) deviates, the more the controllability deteriorates, and the function of the dead time Smith compensation method is not fulfilled, which has a problem that the controllability of the plant is greatly affected.

The present invention has been made in view of the above circumstances, and is capable of fully exhibiting the function of the dead time Smith compensation method against a change in the gain of the controlled object due to a change in the characteristics of the controlled object or a change in the environment. An object of the present invention is to provide a dead time compensation control device capable of executing control with high controllability.

[0011]

According to a first aspect of the present invention, a PI or PID adjustment operation is performed based on a deviation between a target value and a control amount from a control object. to the control system for applying an operation signal that is the control target, the dead time compensation control unit which is obtained by adding the dead time compensation unit according to the dead time Smith compensation method, the
Control amount from control target and output of control target model means
Calculate the corrected gain ratio from
And a model gain correction unit for correcting the output of the dead time compensation unit .

Next, a second aspect of the present invention is a model gain correcting section according to the first aspect of the present invention, wherein a corrected gain ratio is obtained by using a control amount from the controlled object and an output of the controlled object model means. And a multiplier for correcting the output evaluation of the dead time compensator by multiplying the output of the dead time compensator by the corrected gain ratio obtained by the corrected gain ratio acquirer.
It is a configuration with steps and

According to a third aspect of the present invention, there is provided the correction gain ratio acquiring means according to the second aspect of the present invention, wherein the controlled variable from the controlled object is divided by an output of the controlled object model means to obtain a corrected gain ratio. And a time average value calculating means for obtaining a time average value of the output obtained by the division means and setting the corrected gain ratio as a time average value.

[0014]

Therefore, according to the present invention, when the gain of the controlled object changes due to a characteristic change, an environmental change, or the like, the correction gain ratio is determined in accordance with the change in the controlled object gain. Since the evaluation of the output of the time compensating unit is automatically corrected, the gain of the control target model matches the gain of the control target, the condition of the dead time Smith compensation method is maintained, and controllability can be improved.

[0015]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to FIG. 2, the same functions or portions as those in FIG. 2 are denoted by the same reference numerals, and detailed description thereof will be omitted.

That is, in the apparatus of the present invention, the dead time compensating section 5 is provided with the model gain correcting section 10 for automatically correcting the gain of the dead time compensating section 5 in response to the gain change of the control target 2. It is in.

The model gain correction unit 10 takes in the control amount PV _n from the control target 2 and the output signal PV _Mn of the control target model means 6 which constitutes a part of the dead time compensating unit 5, and obtains the ratio k. a dividing means 11 for calculating a _n = PV _n / PV _Mn, the temporal average value of the output k _n of the divider means 11 calculates the average ratio k _n '(hereinafter, for convenience, a dash'
, Which represents the average), and automatically evaluates the output of the dead time compensator 5 by multiplying the output of the dead time compensator 5 by the average ratio k _n ′. And a multiplication means 13 for correction.

[0018] Thus, in this model gain correction unit 10, divides the output signal PV _Mn of the controlled object model means 6 constituting a part of the controlled variable PV _n and delays compensator 5 from the control object 2 means 11 Where PV _n / PV _Mn
After performing the following calculation to obtain the ratio k _n , it is sent to the subsequent time average value calculation means 12. In the temporal average value calculating means 12, the multiplication performed calculation of the temporal mean value of the ratio k _n from the division means 11 calculates the average ratio k _n 'means 13
And the output of the dead time compensator 5 is added to the average ratio k
By multiplying _n ', the evaluation of the output signal of the dead time compensator 5 is automatically changed.

Next, the correction of the controlled object model gain with respect to the change of the controlled object gain will be described. Now, if the characteristics of the controlled object 2 and the controlled object model 7 have done so, it should for the controlled variable PV _n output signals PV _Mn of the controlled object model 7 are the same, the ratio k _n at this time = PV _n / P
If V _Mn , then k _n = 1. However, in practice, k _n ≠ 1 due to the deviation of the characteristics of the control target 2. Therefore,
In this apparatus, the change of the control target gain, i.e. to detect the ratio k _n, is intended to correct the deviation of the controlled object model gain according to the ratio k _n. Now, the output signal PV of the controlled variable PV _n and the controlled object model 7 to the operation signal MV _{n
Mn is, PV n = MV n · {} K P / (1 + T P · s)} · e -LP ^s ... (3) PV _Mn = MV _n · {K _M / (1 + T _M · s)} · e ^{-LP s} (4) can be expressed. Here, the ratio k _n is the output of the divider means 12, since a _{k n = k n = PV n} / PV Mn ...... (5), this (5) (3) and (4) by _{substituting, k n = (K P /} K M) · {(1 + T M · s) / (1 + T P · s)} · e - (LP-LM) s ... (6) here,

[0020]

(Equation 1) Is obtained.

Furthermore, in order to eliminate the influence of momentary fluctuations of ratio k _n, introducing the ratio k _n in time average value calculating means 12, here taking the temporal mean value of the predetermined time, the output 'If _{you, k n' k n = (} K P '/ K M') becomes a ... (8). Accordingly, it leads to the average ratio k _n 'to the multiplying means 13, if Jozure the output of dead time compensator 5, output M _on the said multiplication means 13,

[0022]

(Equation 2)

Is obtained. That is, by adding the model gain correction unit 10 to the output side of the dead time compensating unit 5, the gain of the control target model 7 of the dead time compensating unit 5 is always K _p , that is, as is clear from the equation (9). The gain of the control target 2 is automatically corrected.

Therefore, according to the configuration of the embodiment as described above, the controlled object model 7 is controlled in accordance with the change of the controlled object gain.
By automatically correcting the gain of the control target, it is possible to always match the control target characteristic with the control target model 7, and the dead time Smith compensation method can be applied to the change in the characteristic of the control target and the environmental change. And controllability can be improved. In addition, the control amount from the control target 2 can be divided by the output of the control target model 7 to obtain the correction gain ratio, and the gain of the control target model 7 can be corrected with a very simple configuration.

Further, the ratio obtained by dividing the control amount from the control target 2 by the output of the control target model 7 is set as a temporal average value for a predetermined time by the temporal average value calculating means 12, whereby the ratio The control target model gain can be accurately corrected without being affected by the instantaneous fluctuation.

In particular, in an actual plant, the characteristics of the controlled object change greatly and frequently, but the gain of the controlled object model 7 can be automatically corrected to match the fluctuating gain of the controlled object. Even a control system having time can greatly contribute to improvement of controllability.

[0027] In the above embodiment, although the ratio k _n obtained by the division means 12 by the time average value calculating means 12 to obtain a time average of the predetermined time, the temporal average value calculating means 1
Even if 2 is deleted, an approximate effect can be obtained. In addition, the present invention can be implemented with various modifications without departing from the scope of the invention.

[0028]

As described above, according to the present invention, it is possible to make full use of the function of the dead time Smith compensation method even for a change in the gain of the controlled object due to a change in the characteristics of the controlled object or a change in the environment. In addition, it is possible to provide a dead time compensation control device that can execute control with high controllability even in a control system having a dead time.

[Brief description of the drawings]

FIG. 1 is a functional block diagram showing an embodiment of a dead time compensation control device according to the present invention.

FIG. 2 is a functional block diagram of a conventional dead time compensation control device.

[Explanation of symbols]

1 ... deviation calculating means, 2 ... controlled object, 3 ... PI or PI
D adjustment means, 4 ... subtraction means, 5 ... dead time compensation unit, 6 ...
Primary delay model means, 7: controlled object model means, 8: subtraction means, 10: model gain correction unit, 11: division means,
12 ... time average value calculation means, 13 ... multiplication means.

──────────────────────────────────────────────────続 き Continued on the front page (58) Field surveyed (Int.Cl. ⁶ , DB name) G05B 13/04 G05B 11/36 501 G05B 11/36 503 G05B 11/36 505 G05B 11/36 507

Claims (3)

(57) [Claims] 1. A PI or PID (P: proportional, I: integral, PID) based on a deviation between a target value and a control amount from a control target.
D: derivative) adjusting operation is executed, against the operation signal obtained in the control system to be applied to the controlled object, dead time and first-order lag and the control object model means approximated by a dead time from the control object model unit A dead time compensating control device including a dead time compensating unit based on a dead time Smith compensation method having first-order delay model means for eliminating the control amount, the control amount from the control target and the control target model means
Calculate the corrected gain ratio from the output and this corrected gain ratio
A dead time compensation control device, comprising a model gain correction unit that corrects the output of the dead time compensation unit by using . 2. A corrected gain ratio obtaining means for obtaining a corrected gain ratio using a control amount from the controlled object and an output of the controlled object model means, and a corrected gain ratio obtaining means for obtaining the corrected gain ratio. 2. The dead time compensation control device according to claim 1 , further comprising: multiplying means for correcting the output evaluation of the dead time compensator by multiplying the output of the dead time compensator by the corrected gain ratio. 3. A correction gain obtaining means for dividing a control amount from the control target by an output of the control target model means to obtain a correction gain ratio, and a temporal gain of the output obtained by the division means. JP that has a temporal average value calculating means and the correction gain ratios the average value
The dead time compensation control device according to claim 2 , wherein: